The field of the disclosure relates generally to data transmission systems, and more particularly, to attenuation testing on data transmission lines.
In conventional data transmission systems, two impedance mismatches can create an echo tunnel on a transmission line, such as a cable line. A receiver of a conventional system will then observe a resulting ripple in the frequency response of the transmission line, as well as an impulse in the time domain response, which is echoes are normally delayed relative to a main impulse. However, when only one reflection is present, a resulting frequency response will be flat at the receiver (i.e., not rippled), but have a relatively lower amplitude due to signal loss. It is therefore difficult, in conventional data transmission systems, to discover line defects, such as loose fittings and radial cracks, where an echo tunnel is not formed because only one impedance mismatch is encountered. This problem is rendered more difficult due to the fact that time domain reflectometer (TDR) test equipment cannot be used on in-service cable plant. Accordingly, there is a need to be able to locate line defects on an in-service data transmission line when only one impedance mismatch is observed.
Additionally, conventional cable operators experience a problem with aging cable lines that, over time, experience a variety of faults. A shield break in the cable line, for example, is a fault producing a discrete reflection from one point in the line. Other faults, such as water seeping into the cable, will increase signal attenuation through the cable, even if the seepage only sometimes produces reflections. One conventional solution utilizes adaptive equalizers to compensate for reflections that are not too significant, but this solution does not address situations where the reflections are severe. Moreover, flat signal attenuation experiences an additional noise problem due to low signal level pushing the desired signal into a noise floor.
Another conventional solution utilizes a network analyzer to measure a length of coaxial cable, but only when both ends of the measured length are in the same location. When the ends are not located together, a transmitter may be placed at one end of the measured length, and a receiver at the other end. A training signal is transmitted from end to end, and channel response phase information and magnitude are computed for the captured training signal. This solution is not implemented on an in-service line.
When an in-service cable line experiences transmission difficulties, the questionable cable line must be assessed for the line quality. That is, the line must be checked for faults due to problems such as water damage, stress fractures, corrosion, bad connectors, animal chews, or other mechanical damage. The line may have excessive attenuation, reflections, or both. The line must also be checked to determine if there has been an attempt to cut into the line for signal sharing, such as from an illegal tap. Conventional testing schemes measure the broadband signals at both ends of the questionable line, and subtract the difference for attenuation versus frequency to estimate the source of the fault. This conventional solution also inconveniently requires active measurement equipment at both ends of the line.